LinkedIn connection request limits: what is actually known, and what is guesswork

Search for LinkedIn’s connection request limit and you’ll be told 100 a week. Or 200. Or 20 a day, or 30, or “up to 100 depending on your account”. Every one of those numbers is stated with total confidence, and they cannot all be right.

They can all be wrong, though, and mostly they are — not because the people writing them are careless, but because they are answering a question LinkedIn has never answered. Here is what is actually known, what is inference dressed up as fact, and what to do given that the distinction matters more than the number.

What LinkedIn has actually said

LinkedIn introduced a weekly invitation limit in 2021 and confirmed publicly that one exists. It has never published the figure, and it has never committed to a single figure applying to every account.

What LinkedIn’s own help pages do say, in substance:

• There is a limit on how many invitations you can send, and it is applied weekly
• The limit is not the same for every account
• Accounts with a high proportion of ignored or withdrawn invitations may see theirs reduced
• Sending invitations to people who then report them as unwanted affects standing directly

That last point is the one most articles skip, and it matters more than the arithmetic. LinkedIn is not primarily counting your invitations. It is watching what happens to them.

What is inference, not fact

The specific figures circulating — 100 a week, 200 for some accounts, 20 a day as a “safe” rate — are observations, not documentation. They come from people watching their own accounts hit a wall and reporting where the wall was.

That’s genuinely useful information. It is not a specification. Three things make it unreliable as one:

It varies by account. Two accounts sending the same volume in the same week get different outcomes depending on tenure, network size, past acceptance rate and prior warnings. Someone reporting “the limit is 100” is reporting their limit, that week.

It changes. LinkedIn adjusts enforcement. A number that was right in 2023 may not be right now, and nothing announces the change except accounts hitting it.

Reporting is biased. The people who write about the limit are disproportionately the ones who hit it. Accounts quietly sending 15 a day for two years do not write blog posts about a limit they never reached.

Why “the limit” is the wrong thing to optimise

Suppose you did know the exact number. You would then face a more awkward question: is running right up against it a good idea?

Almost never. Sitting at 95% of a threshold you cannot see, on a system that adjusts that threshold based on how recipients react to you, is not a stable position. One bad week — a poorly targeted batch, a few reports — and you are on the wrong side of a line you were already touching.

There is also a quieter cost. A rate that requires you to be at the limit implies a targeting problem. If 100 invitations a week produce three conversations, the problem is not that you cannot send 150.

What the acceptance rate actually tells you

The single most useful number in LinkedIn outreach is not how many invitations you sent. It is what share of them were accepted.

A high acceptance rate means you are contacting people who recognise why you are contacting them. It keeps the account in good standing, and it is the strongest available signal that your targeting is sound.

A low one means the opposite, and it compounds: LinkedIn sees ignored invitations pile up, your standing drops, your effective limit drops with it, and the response is usually to send more — which makes all three worse.

If you change one thing after reading this, make it the thing you measure. Sent volume is an input. Acceptance rate is a result.

A working rate, and where it comes from

Given that nobody can tell you the real threshold, the honest approach is to pick a rate well under any number anyone has reported, and stay there.

That is what the limits Dizenly enforces are built on: 25 connection requests a day, 125 a week, spread across working hours with randomised gaps rather than fired in a burst. New accounts start at a third of that and build over two weeks. The limits can be set lower and cannot be raised.

Those are not LinkedIn’s numbers — nobody has LinkedIn’s numbers. They are numbers chosen to sit comfortably below anything observed, on the reasoning that the cost of being slightly too cautious is a few fewer invitations, and the cost of being slightly too aggressive is the account.

You can work out a rate for a specific account from its tenure, network size and history — the arithmetic is shown on screen so you can disagree with it.

The practical version

If you want four rules rather than an essay:

1. Stay well under any figure you have read. The gap between 25 a day and whatever the real ceiling is, is your safety margin. Do not spend it.
2. Watch acceptance rate, not sent volume. Below about 30%, stop and fix the targeting before sending anything else.
3. Withdraw invitations nobody answered. Pending invitations that sit for weeks count against you. Clearing them after a week or so is free.
4. Ramp new accounts slowly. An account that goes from quiet to busy is the single most recognisable pattern there is.

None of this is exciting, and none of it will let you send a thousand invitations a week. It will let the account still be there next year, which turns out to be the constraint that matters.

The part nobody can promise

No tool, including ours, can guarantee an account will never be restricted, and any vendor who tells you otherwise is guessing on your behalf. LinkedIn changes enforcement without notice, and none of this is legal advice — read the current User Agreement yourself rather than take a vendor’s word for it.

What you can control is the margin you leave, and whether the people you contact are glad you did. What actually triggers a restriction covers the rest of the enforcement picture, including what to do if you are already on the wrong side of it.